Are liquidated damages clauses enforceable in Texas?

Texas favors freedom of contract, as a policy "firmly embedded in our jurisprudence."¹⁴ But tempering this policy is the "universal rule" that damages for breach of contract are limited to "just compensation for the loss or damage actually sustained." Stewart v. Basey, 150 Tex. 666, 245 S.W.2d 484, 486 (1952). Accordingly, courts carefully review liquidated damages provisions to ensure that they "adhere to the principle of just compensation."¹⁵

In keeping with this approach, an enforceable liquidated damages contract provision establishes an "acceptable measure of damages that parties stipulate in advance will be assessed in the event of a contract breach."¹⁶ A damages provision that violates the rule of just compensation, however, and functions as a penalty, is unenforceable.¹⁷ Liquidated damages must not be punitive, neither in design nor operation.¹⁸

We first considered the difference between an enforceable liquidated damages provision and an unenforceable penalty in Stewart v. Basey, 150 Tex. 666, 245 S.W.2d 484, 485-486 (1952). There we explained:

Volumes have been written on the question of when a stipulated damage provision of a contract should be enforced as liquidated damages and when enforcement should be denied because it is a penalty provision.... All agree that to be enforceable as liquidated damages the damages must be uncertain and the stipulation must be reasonable.

. . . . .

The right of competent parties to make their own bargains is not unlimited. The universal rule for measuring damages for the breach of a contract is just compensation for the loss or damage actually sustained. By the operation of that rule a party generally should be awarded neither less nor more than his actual damages. A party has no right to have a court enforce a stipulation which violates the principle underlying that rule.

More recently, in Rio Grande Valley Sugar Growers, Inc. v. Campesi, 592 S.W.2d 340, 342 n. 2 (Tex.1979), we restated the two-part Stewart test for determining whether to enforce a contractual damages provision as follows: "In order to enforce a liquidated damage clause, the court must find: (1) that the harm caused by the breach is incapable or difficult of estimation, and (2) that the amount of liquidated damages called for is a reasonable forecast of just compensation." Cf. TEX.BUS. & COM.CODE § 2.718(a). ³

Whether a contractual provision is an enforceable liquidated damages provision or an unenforceable penalty is a question of law for the court to decide. Farrar v. Beeman, 63 Tex. 175, 181 (1885); see Lefevere v. Sears, 629 S.W.2d 768, 771 (Tex.Civ.App.--El Paso 1981, no writ); Muller v. Light, 538 S.W.2d 487, 488 (Tex.Civ.App.--Austin 1976, writ ref'd n.r.e.); Schepps v. American Dist. Telegraph Co., 286 S.W.2d 684, 690 (Tex.Civ.App.--Dallas 1955, no writ); Zucht v. Stewart Title Guar. Co., 207 S.W.2d 414, 418 (Tex.Civ.App.--San Antonio 1947, writ dism'd); Bourland v. Huffhines, 244 S.W. 847, 849 (Tex.Civ.App.--Amarillo 1922, writ dism'd). Sometimes, however, factual issues must be resolved before the legal question can be decided. For example, to show that a liquidated damages provision is unreasonable because the actual damages incurred were much less than the amount contracted for, a defendant may be required to prove what the actual damages were. See Johnson Eng'rs, Inc. v. Tri-Water Supply Corp., 582 S.W.2d 555, 557 (Tex.Civ.App.--Texarkana 1979, no writ); Oetting v. Flake Uniform & Linen Serv., Inc., 553 S.W.2d 793, 796-797 (Tex.Civ.App.--Fort Worth 1977, no writ); Smith v. Lane, 236 S.W.2d 214, 215 (Tex.Civ.App.--San Antonio 1950,

Page 789

no writ); Southern Plow Co. v. Dunlap Hardware Co., 236 S.W. 765, 766-767 (Tex.Civ.App.--Dallas 1922, no writ); Walsh v. Methodist Episcopal Church, 212 S.W. 950, 952 (Tex.Comm'n App.1919, judgm't adopted).

In Phillips v. Phillips, we emphasized that courts will enforce liquidated damages provisions when: (1) "the harm caused by the breach is incapable or difficult of estimation," and (2) "the amount of liquidated damages called for is a reasonable forecast of just compensation."¹⁹ In applying the first two rules, courts examine the circumstances at the time the agreement is made.²⁰ The party seeking liquidated damages bears the burden of showing that the provision, as drafted, accounts for these two considerations.²¹

A properly designed liquidated damages provision, however, may still operate as a

[595 S.W.3d 193]

penalty due to unanticipated events arising during the life of a contract. Thus, we observed in Phillips that courts must also examine whether "the actual damages incurred were much less" than the liquidated damages imposed, measured at the time of the breach.²²

When a contract’s damages estimate proves inaccurate, and a significant difference exists between actual and liquidated damages, a court must not enforce the provision. Applying this rule in FPL Energy, LLC v. TXU Portfolio Mgmt. Co., we held that the "unacceptable disparity" between damages assessed under the contract (approximately $29 million) and actual damages (approximately $6 million) made the liquidated damages provision unenforceable.²³ At the time of contracting, damages from a breach in that case "were difficult to estimate" and the liquidated damages provision "on [its] face, reasonably forecast damages."²⁴ Nonetheless, we held the provision unenforceable because it "operate[d] with no rational relationship to actual damages."²⁵ When an "unbridgeable discrepancy" exists between "liquidated damages provisions as written and the unfortunate reality in application," the provisions are not enforceable.²⁶ To avail itself of this defense, the breaching party challenging a provision must demonstrate this unbridgeable discrepancy.²⁷

Phillips did not create a broad power to retroactively invalidate liquidated damages provisions that appear reasonable as written. See820 S.W.2d at 788. Nor do we create such a power here. But when there is an unbridgeable discrepancy between liquidated damages provisions as written and the unfortunate reality in application, we cannot enforce such provisions. The forecast of damages was flawed by its reliance on events that did not and perhaps cannot occur—a PUC determination of the market value of RECs and a failure to secure a regulatory excuse for curtailment-based REC deficiencies. When the liquidated damages provisions operate with no rational relationship to actual damages, thus rendering the provisions unreasonable in light of actual damages, they are unenforceable. See id. Because the liquidated damages provisions operate as a penalty, we hold the provisions unenforceable.

The enforceability of the contractual provision in this case involves no fact issues. A contractual provision like the one here by which one party agrees to pay the other some multiple of actual damages for breach of the agreement does not meet either part of the legal test for an enforceable liquidated damages provision. It cannot meet the first prong of the test because the harm caused by the breach of the contract is not incapable or difficult of estimation. The provision assumes actual damages can and will be determined, indeed must be determined, before the prescribed multiplier can be applied. The provision cannot meet the second prong of the test because, instead of attempting to forecast actual damages, it calls for them to be determined and then multiplied. Cf. Robert G. Beneke & Co. v. Cole, 550 S.W.2d 321 (Tex.Civ.App.--Dallas 1977, no writ) (contract provision which fixes liquidated damages without excluding additional liability for actual damages is not a reasonable forecast of just compensation and therefore a penalty). A contractual provision like the one in this case is thus, on its face, an unenforceable penalty.

Ryan Steen, president of ImageFirst, testified that, at the time the parties made their agreement, it was difficult to estimate the damages that would result if Atrium canceled the contract. At the time of trial, Steen had worked for ImageFirst for 11 years and had performed under approximately 120 contracts in two markets, including Houston. He explained that a customer’s "burn rate" of linens—the percent of linens that must be discarded per cycle because they are stained or otherwise unusable—is unknown and difficult to calculate at the outset. And a customer’s patient load can fluctuate, making Atrium’s

[595 S.W.3d 194]

weekly demand for linens unpredictable. All this, in turn, affected ImageFirst’s ability to forecast its weekly gross margin.

As it turned out, Atrium’s patient load increased in the first months of the contract. ImageFirst escalated its deliveries to Atrium from two or three days a week to five or seven days a week. Corresponding to this increased usage, Atrium’s weekly invoices increased from approximately $2,600 (when the contract was signed) to between approximately $7,000–$8,000 in later weeks. Steen stated that these variations are "typical in the industry." Steen’s testimony supports the trial court’s finding that damages for breach of the contract were difficult to predict at the time of contracting.

Atrium presented no evidence rebutting Steen’s testimony. Instead, Atrium challenged ImageFirst’s interpretation of the cancellation provision, contending that the contract limited ImageFirst to its reliance damages, that is, the cost of linens and supplies purchased in connection with the contract. Relying on ImageFirst’s invoices, Atrium observed that this reimbursement cost is not difficult to estimate. Atrium asserts that these damages are determinable, and significantly less, than the amount provided for in the liquidated damages provision. We reject Atrium’s challenge for two reasons.

First, the contract did not limit ImageFirst’s damages to reimbursement of its used linens and supplies. Atrium urges this reading by noting that the liquidated damages were intended to "reimburse[ ] [the] Company for related investments to service the customer." The cancellation provision, however, contemplates recompense for ImageFirst’s lost benefit of the bargain—or expectancy—damages. It defines "reimbursement" for the loss of ImageFirst’s "investment" as equal to a percentage of the "agreement value" or most recent invoice, multiplied by the remaining weeks of the contract. This measure does not limit ImageFirst to reimbursement for its out-of-pocket costs.

Second, difficulty of estimation is evaluated based on the circumstances at the time the agreement is made. In this requirements contract, the parties expected demand would vary, and thus they could not know the average weekly gross margin at the time they made the agreement. Atrium relies on the actual invoices, which eventually varied little. Because courts assess the difficulty of estimation from the perspective of the parties at the time of contracting, however, Atrium’s argument that damages became calculable at the time of the breach is unavailing.³⁰ The significant difference between the initial invoice and those preceding cancellation indicates that damages were hard to estimate until laundry services were under way.

Atrium next challenges the trial court’s finding that the contract’s estimate of damages was a reasonable forecast of just compensation. ImageFirst, however, presented evidence that the gross margin amount was based on the franchisor’s "historical profit margin" from over forty years in the industry. Steen also examined "the books and records" of ImageFirst to determine "if that 40 percent [was] about right" for ImageFirst in particular. He

[595 S.W.3d 195]

found that it was: "[G]oing back to 2012, our [profit] margin, with the exception of 2013, hovered between 32 and 39 percent.... 2013 was an outlier in that it was only 9.2 [percent] and that was almost certainly because of the outstanding balance from Atrium."

Atrium responds that ImageFirst used this agreement for all of its franchises, without regard to local variations in profitability, so the 40 percent number is per se unreasonable.31 ImageFirst’s use of a form business contract is not unreasonable if other evidence demonstrates the forecast’s reasonableness for these parties. In this case, the evidence did. Because Steen explained that 40 percent reflected a reasonable estimate for ImageFirst’s margin based on its actual historical performance, the trial court did not err in finding it to be a reasonable forecast.³²

On appeal, Atrium refers to publications citing gross margins in mostly unrelated industries. But margins in other industries are of little relevance and, importantly, Atrium did not present these publications to the trial court. We do not consider them. Atrium did not counter the reasonableness of a 40 percent gross margin on a laundry-services contract with evidence at trial. And Steen also testified, without contradiction, that a five-year contract term is the industry standard. Whether a liquidated damages provision is a penalty is a legal question, but its resolution may depend on underlying facts.³³ The evidence supports the trial court’s resolution here.

Atrium further responds that the provision is an unreasonable forecast because the weekly invoice used to calculate damages could be above-average. Liquidated damages allow parties to "allocate the risk of uncertainty over the actual loss."³⁴ The parties are expected to negotiate a reasonable—not perfect—forecast of just compensation. Here, it was reasonable to choose the greater of the first or last invoice, given that demand for the services was under Atrium’s control, and later invoices could more accurately reflect losses over the life of the remaining contract if it were prematurely canceled. As we discuss separately, there is a backstop against a forecast that is inordinate when compared with actual damages: proof of a large variance will render a provision unenforceable.

The facts in this case differ from those cases in which a contract facially imposed amounts beyond just compensation. In Phillips, the liquidated damages provision required the breaching party to pay a multiple of actual damages—as

[595 S.W.3d 196]

drafted, this violated the first two Phillips rules.³⁵ And provisions that use the same damage measure for breaches of varying magnitude are also facially unreasonable.³⁶ The contract provision in this case neither multiplies actual damages nor penalizes dissimilar breaches with the same broad brush.

We conclude that the record supports the trial court’s findings that, at the time of contracting, (1) damages resulting from Atrium’s breach were difficult to estimate and (2) the liquidated damages provision reasonably forecast just compensation.

Because ImageFirst carried its burden, we turn to whether Atrium met its burden to show that the liquidated damages provision was otherwise unreasonable when compared with ImageFirst’s actual damages.³⁷ While the court of appeals correctly observed that liquidated damages must compensate "for losses sustained and no more," it did not examine whether an "unbridgeable discrepancy" existed between actual and liquidated damages at the time Atrium canceled the contract.³⁸ The evidence does not demonstrate such a discrepancy, however, so the outcome remains the same.

As we observed in Phillips, "to show that a liquidated damages provision is unreasonable because the actual damages incurred were much less than the amount contracted for, a defendant may be required to prove what the actual damages were."³⁹ Atrium offered no evidence that an unbridgeable discrepancy existed between ImageFirst’s actual expectancy damages and its liquidated damages under the contract. Rather, Atrium contended that ImageFirst’s reliance damages were much less. As Atrium’s counsel told the trial court:

[I]n breach of contract cases, there are two measures of damages that are equally applicable. There’s the expectation, lost profits, in other words; or

[595 S.W.3d 197]

there’s a reliance interest which is basically putting the non-breaching party in the same place that the non-breaching party was [in] before the contract was executed.

Now, by stating expressly that this cancellation fee was a reimbursement to the company ... this contract made an election that it was going to try to use a liquidated damages provision as a substitute for reliance damages.

We have rejected Atrium’s argument that the contract limited ImageFirst’s damages to a reliance measure. And Atrium did not adduce evidence that the liquidated damages provision failed to approximate ImageFirst’s damages for lost benefit of its bargain. Thus, the trial court had no basis to conclude that the liquidated damages were out of step with actual damages.

Moreover, Atrium adduced no evidence that mitigation would have reduced ImageFirst’s actual damages such that liquidated damages no longer served as just compensation.⁴⁰ Courts must consider mitigation when determining whether actual damages diverge significantly from liquidated damages.⁴¹ The burden remains on the party seeking to prove that the liquidated damages provision is an unenforceable penalty to show how much actual damages were reduced, or could have been reduced, by mitigation.

Atrium points to testimony that ImageFirst repurposed Atrium’s linens to service other contracts to support its mitigation argument. Although Steen acknowledged that ImageFirst "repurposed" some of Atrium’s linens after Atrium canceled the contract, Atrium did not attempt to value those linens or to prove that repurposing them materially reduced ImageFirst’s expectancy damages over the remaining life of the contract. When an injured party "would have entered into both transactions but for the breach, [it] has ‘lost volume’ as a result of the breach."⁴² Atrium provided no evidence that ImageFirst replaced Atrium as a customer and could handle no others. Because Atrium did not attempt to quantify either ImageFirst’s expectancy damages or the effect of mitigation, it did not demonstrate that an "unbridgeable discrepancy" exists between ImageFirst’s actual and liquidated damages.

A breaching party need not prove the other party’s actual damages to invalidate penalty provisions in every case— Phillips and Stewart demonstrate the opposite. The liquidated damages provisions in those cases were facially invalid, without extrinsic

[595 S.W.3d 198]

evidence of actual damages, as the measures at the outset could not reasonably forecast actual damages. But when a liquidated damages provision is facially reasonable, the breaching party must present evidence from which the court may find that an "unbridgeable discrepancy" exists between actual and liquidated damages.⁴³

* * *

We hold that, at the time the parties’ agreement was made, (1) the harm that would result from a breach was difficult to estimate and (2) the liquidated damages provision reasonably forecast just compensation. We further hold that the breaching party failed to demonstrate an "unbridgeable discrepancy" between liquidated and actual damages, measured at the time of the breach, to invalidate an otherwise valid contract provision. Accordingly, we affirm the judgment of the court of appeals.

Yet the facts of this case demonstrate the chasm between the liquidated damages provisions as written and the result of the provisions under the court of appeals' judgment. First, the number of deficient RECs varies significantly between TXUPM's assertion and what the regulatory scheme would indicate. FPL had a collective deficiency of 580,000 RECs, yet 62% (or about 360,000) were not produced because of transmission congestion and associated ERCOT curtailment orders, which are excused by the PUC rules. Id. TXU Electric was subject to PUC penalties for REC deficiencies at the time of contract formation. Id. (former 16 Tex. Admin. Code § 25.173(c)(1)). Upon assignment to TXUPM, a power marketer, no party was subject to PUC penalties. See16 Tex. Admin. Code § 25.5(83) (defining power marketer as an wholesaler seller of electricity who does not own generation, transmission, or distribution facilities in Texas, which would exclude TXUPM from REC penalties). This change in relationship did

[426 S.W.3d 72]

not undermine each contract, but it fundamentally changed the basis for the liquidated damages provisions. Those provisions presuppose that TXU Electric or its successors would respond to potential penalties for REC deficiencies. When those successors have no REC penalty obligations, they may, as occurred here, fail to secure a regulatory excuse for deficiencies that would obviate any need for the liquidated damages provisions. If the PUC could assess a penalty against TXUPM, the penalty would be based on the 220,000 RECs attributable to lack of wind, not congestion.

Second, the Deficiency Rate calculation failed to tie the damages to market value as the contracts contemplate. Section 4.04(f) of the contracts allows for a Deficiency Rate of either $50 or twice the annual average market value of RECs “[t]o the extent that the [PUC] determines the annual average market value.” The PUC expressly declined TXUPM's request for such a determination. The actual market value of a REC during the period in question ranged from $4 to $14. The fortuity of a PUC determination thus utterly controls the damages, irrespective of the actual market value. For instance, the appropriate amount of damages should fall in the range of $8 to $28 (twice the average market value), depending on what the PUC would have determined as the actual market value of a REC in each year.

In combination, this creates an unacceptable disparity. The court of appeals assessed damages at $29 million. If we use the REC deficiency of 220,000 (as reduced under PUC rules), and the reduced Deficiency Rate of $28 (the upper bound of the possible range), actual damages equal only $6,160,000. To reach damages of $29 million on a 220,000 REC deficiency would require an effective deficiency rate of $132 per REC. The disparity grows if we consider that TXUPM also avoided the contract price of $24 per MWh of Renewable Energy—which includes a REC and a MWh of electricity. Although only a portion of the $24 is attributable to the REC not purchased, it nonetheless would further diminish TXUPM's actual damages. In Phillips, we recognized that a liquidated damages provision may be unreasonable in light of actual damages. 820 S.W.2d at 788. The burden of proving unreasonableness falls to FPL. See id. The court of appeals held that FPL failed to meet this burden, yet the court's evaluation was based on evidence of damages for electricity and RECs. 328 S.W.3d at 589–90. Our holding on the scope of the liquidated damages clauses limits our consideration to damages for REC deficiencies. The evidence reviewed in this opinion demonstrates that FPL has met its burden.

Phillips did not create a broad power to retroactively invalidate liquidated damages provisions that appear reasonable as written. See820 S.W.2d at 788. Nor do we create such a power here. But when there is an unbridgeable discrepancy between liquidated damages provisions as written and the unfortunate reality in application, we cannot enforce such provisions. The forecast of damages was flawed by its reliance on events that did not and perhaps cannot occur—a PUC determination of the market value of RECs and a failure to secure a regulatory excuse for curtailment-based REC deficiencies. When the liquidated damages provisions operate with no rational relationship to actual damages, thus rendering the provisions unreasonable in light of actual damages, they are unenforceable. See id. Because the liquidated damages provisions operate as a penalty, we hold the provisions unenforceable.